Reviewing
One crucial aspect of being a scientist is peer-review of both papers and grants from your colleagues. We will have internal journal reviewing debates and practice proposal view sessions, but from time to time, I will ask you to help me review a manuscript that is under consideration for publication at a journal. Here are some guidelines of what to keep in mind while reviewing papers:
Excellent eLife Ambassador's Guide to Peer Review: a reading list to help you "identify and address common problems with published articles" (re-created in part below):
1. Why you shouldn't use bar graphs to show continuous data (and what to do instead)
described here, and addressed here & here & here
2. The problem with underpowered studies (Low power is a problem even when you find a significant difference)
described here
Small samples are more likely to give spurious results: described here
3. Why it's important to report all excluded observations & the reasons for their exclusion
described here
4. P-values are often reported incorrectly: Why it's important to present the information needed to verify the test result
described here
Common misconceptions about data analysis and statistics: described here
Research methods: know when your numbers are significant: described here
5. Unblinded studies find larger effects
described here
6. Animal research: Follow the ARRIVE guidelines to improve transparency and reproducibility
7. Animal research: Multi-lab studies may improve reproducibility
described here
8. Check for clusters of non-independent data (replicates, mice from the same litter, correlated variables, etc.): Did the authors account for non-independence in their analysis?
described here
9. Beware of image manipulation (and plotting that hides data or misrepresents results)
described here
Transparency is the key to quality: described here
[I'll be updating this section shortly, and constantly as additional items come up]